1
2
3
4
5
6
7
8
9

<VirtualHost *:80>
    ServerName lab.sgmac.com
    Redirect / https://lab.sgmac.com/
</VirtualHost >

<VirtualHost *:443>
    ServerName lab.sgmac.com
    ... 
</VirtualHost >

I find this pretty straightforward and more readable.

These are my last two commits on a repository with the wrong email address,

1
2
3
4
5
6
7
8
9
10
11

commit 2c096fc8c8fa425bfd6c11a1223ec9a512383784
Author: Srgio Galvan <false_email@gmail.com>
Date:   Thu Jun 28 15:40:08 2012 +0300

    Add filter example

commit b70b8832df12a53069ae36e928ffb7fbce2dea2e
Author: Sergio Galvan <false_email@gmail.com>
Date:   Wed Jun 27 00:22:18 2012 +0300

    Last exercise

Then I found git filter-branch(1) :

1

git filter-branch --env-filter ' export GIT_AUTHOR_EMAIL=good_email_address@gmail.com'

It’s pretty straighforward, let’s check the new email address:

1
2
3
4
5
6
7
8
9
10
11

commit 4cf11c1d39fd4d91792d37f90ca51eb5b9037477
Author: Sergio Galvan <good_email_addres@gmail.com>
Date:   Thu Jun 28 15:40:08 2012 +0300

    Add filter example

commit f391ad1d131b287df8aa7df15e01ce707b64419f
Author: Sergio Galvan <good_email_addressgmail.com>
Date:   Wed Jun 27 00:22:18 2012 +0300

    Last exercise

As you might already know, the SHA-1 hash is different, now you can’t keep pushing to your remote if you ever had one. On the other hand, I have my respository almost ready to pushing into github. First create your respository in github, and the proceed with the next step:

1

git  push --mirror git@github.com:githubuser/myrepo.git

There are more properties you might want to define, check theme out.

Basically the module allows to define one or more repositories, besides of the area. In Debian we find three areas, main,contrib and non-free, though the last two are not part of Debian. In any case, you could find different areas when using non-official repositories. The module itself only updates the /etc/apt/sources.list file, you do not forget to update your system afterwards.

1
2
3
4
5
6
7
8
9
10
11
12
13

 class apt-sources (
        $release = $::lsbdistcodename,
        $area = undef,
        $sources = undef
){
    file { '/etc/apt/sources.list':
            ensure  => present,
            owner   => 'root',
            group   => 'root',
            mode    => '0644',
            content => template('apt-sources/sources.list.erb');
    }
}

And now we only need to define the template.

1
2
3
4
5
6
7
8
9
10
11
12
13

<% @area = area.is_a?(Array) ? area.flatten.join(" ") : area  %>

# Default repositories
deb http://ftp.ie.debian.org/debian/ <%= release %> main
deb-src http://ftp.ie.debian.org/debian/ <%= release %> main

<% if sources.is_a?(Array) %>
<% sources.each do |repo| -%>
deb <%= repo %> <%= release %> <%= @area %>
<% end -%>
<% else %>
deb <%= sources %> <%= release %> <%= @area %>
<% end %>

And some examples:

1
2
3
4
5
6
7
8
9
10
11
12
13
14

# Example 1: Just one source and one area

class { 'apt-sources':
        area    => 'main',
        sources => 'http://backports.deban.org/debian-backports',
}

# Example 2:  Using arrays

class { 'apt-sources':
        area    => ['main','non-free'],
        sources => ['http://backports.deban.org/debian-backports',
          'http://www.deb-multimedia.org']
}

This approach is quick and useful, and does the work. Just a bunch of ideas to improve it: Keyring management, Management of areas in apt-sources.d, and apt-sources as a puppet type.

First of all, we can convert from VDI to raw. I did the conversion from Windows, I guess it should be the same from Linux.

1

 C:/VDIs/VboxManage internalcommands converttoraw  disk-vm-testing.vdi  vdisk.raw

Now let’s see what is inside the raw disk:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18

dm@testing:#{~} fdisk  -l vdisk.raw
You must set cylinders.
You can do this from the extra functions menu.

Disk vdisk.raw: 0 MB, 0 bytes
255 heads, 63 sectors/track, 0 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x000d598a

    Device Boot      Start         End      Blocks   Id  System
vdisk.raw1   *           1         996     7993344   83  Linux
Partition 1 does not end on cylinder boundary.
vdisk.raw2             996        1045      392193    5  Extended
Partition 2 has different physical/logical endings:
     phys=(1023, 254, 63) logical=(1044, 52, 32)
vdisk.raw5             996        1045      392192   82  Linux swap / Solaris

In this case I only have a first partition, because the second one was used as swap device. I have to find out the offset where the data is.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18

dm@testing:#{~}  parted vdisk.raw
GNU Parted 2.3
Using /media/sf_Downloads/vdisk.raw
Welcome to GNU Parted! Type 'help' to view a list of commands.
(parted) unit
Unit?  [compact]? B
(parted) p
Model:  (file)
Disk /media/sf_Downloads/vdisk.raw: 8589934592B
Sector size (logical/physical): 512B/512B
Partition Table: msdos

Number  Start        End          Size         Type      File system     Flags
 1      1048576B     8186232831B  8185184256B  primary   ext4            boot
 2      8187280384B  8588886015B  401605632B   extended
 5      8187281408B  8588886015B  401604608B   logical   linux-swap(v1)

(parted)q

The ‘Start’ column shows me the offset for the partition I’m interested in. Next step is to map this offset with a loopback device and mount it.

1
2
3
4
5
6
7

dm@testing:#{~} losetup -o 1048576  /dev/loop0  vdisk.raw
dm@testing:#{~} mount /dev/loop0 /mnt/vdi
dm@testing:#{~} mount | grep loop
/dev/loop0 on /mnt/vdi type ext4 (rw)
dm@testing:#{~} ls /mnt/vdi
ls /mnt/
bin  boot  dev  etc  home  initrd.img  lib  lost+found  media  mnt  opt  proc  root  sbin  selinux  srv  sys  tmp  usr  var  vmlinuz

At this point you should be able to access your data, or at least some part of it.

The function below takes a list of directories and creates bare respositories on the remote server.

1
2
3
4
5
6
7
8
9
10
11
12
13
14

#############################
# Create several repostiories
#############################

g() {
   [ $# -lt 1 ] && printf 'repo name missing\n' && return 126 ;
   for((i=1; i <= $#; i++))
   do
        eval repo='$'$i
        ssh git@server.com  " mkdir ${repo}.git && cd ${repo}.git && git init --bare &> /dev/null"
   done

 echo "${repo}.git" >> $HOME/.repolist
}

In order to keep a list of my repositories, every time I create a new repository I append it to my repolist. Sometimes I have been working on something and I want to add the current directory to a remote repository, here is another function that helps:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18

#############################
# Gitfy the current directory
#############################

gitfy() {
   g ${PWD##*/}

cat <<-EOF > .gitignore
*~
*#
EOF

   git init . && git add . &&\
                 git commit -m "First commit" && \
                 git remote add origin ssh://git@server.com/home/git/${PWD##*/}.git && \
                 git push origin master

}

Here is the bash completion for my git repositories :

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22

##############################
# Repositories bash completion
##############################
clone() {
        git clone ssh://git@server.com/home/git/${1}
}

_clone() {

    local cur prev opts
    COMPREPLY=()
    cur="${COMP_WORDS[COMP_CWORD]}"
    prev="${COMP_WORDS[COMP_CWORD-1]}"
    opts=$( cat $HOME/.repolist )

    if [[ ${cur} == * ]] ; then
        COMPREPLY=( $(compgen -W "${opts}" ${cur}) )
        return 0
    fi

}
complete -F _clone clone

And the example :

1
2
3
4
5

adm@testing:${~}clone [TAB]
mygits.git             puppet-template.git    tcp.git
capistrano.git         course-saas.git        dotfiles.git
ldap.git               notes.git              puppet-vps.git
todos.git

I’m quite sure it could be much better but for me works fine. It’s pretty handy, and I had the opportunity to play around with bash completion.

Both of them do the job, the question here, we want to use the best one. Suddenly I remembered an option that comes with strace and lets you query the number of syscalls for a given trace. Besides we can order based on the number of syscalls.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23

adm@testing:${~} strace -c -S calls kill -0 1234
kill: No such process
% time     seconds  usecs/call     calls    errors syscall
------ ----------- ----------- --------- --------- ----------------
  -nan    0.000000           0        12           mmap2
  -nan    0.000000           0         6           close
  -nan    0.000000           0         5           open
  -nan    0.000000           0         5           fstat64
  -nan    0.000000           0         4           read
  -nan    0.000000           0         4         4 access
  -nan    0.000000           0         3           brk
  -nan    0.000000           0         3           munmap
  -nan    0.000000           0         2           mprotect
  -nan    0.000000           0         1           write
  -nan    0.000000           0         1           execve
  -nan    0.000000           0         1           getpid
  -nan    0.000000           0         1         1 kill
  -nan    0.000000           0         1           dup
  -nan    0.000000           0         1         1 _llseek
  -nan    0.000000           0         1           fcntl64
  -nan    0.000000           0         1           set_thread_area
------ ----------- ----------- --------- --------- ----------------
100.00    0.000000                    {52}       6 total

On the other hand, the second command’s output:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

adm@testing:${~}strace -c -S calls test -d /proc/1234
% time     seconds  usecs/call     calls    errors syscall
------ ----------- ----------- --------- --------- ----------------
  -nan    0.000000           0         7           mmap2
  -nan    0.000000           0         5           close
  -nan    0.000000           0         3           open
  -nan    0.000000           0         3         3 access
  -nan    0.000000           0         3           brk
  -nan    0.000000           0         3           fstat64
  -nan    0.000000           0         2           mprotect
  -nan    0.000000           0         1           read
  -nan    0.000000           0         1           execve
  -nan    0.000000           0         1           munmap
  -nan    0.000000           0         1         1 stat64
  -nan    0.000000           0         1           set_thread_area
------ ----------- ----------- --------- --------- ----------------
100.00    0.000000                   {31}         4 total

The column calls helps to know the number of performed syscalls. Using test -d /proc/PID gives a better performance due to a minor number of syscalls.

I really like strace, is a tool you had better know, here I got syscalls statistics, but you can trace syscalls either specific or a bunch of them, follow forked processes and much more, this is only a simple example. I hope it helps.

Basically the script takes a number of directories and an optional pattern (last argument) from the command line. If either a directory or a pattern is not providen, it looks for in the current directory, and use a default pattern.

The script searches for those files matching the pattern and deletes thoses files from the system. Besides, it checks the owner of the files is the same that is running the script.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22

#!/usr/bin/env ruby
require 'find'
directories = []

pattern = /.*~$/
if ! ARGV.empty?  and ARGV.size > 1
  pattern = /#{ARGV.pop}/
end

directories = (ARGV.empty?) ? File.expand_path(File.dirname(__FILE__)) : ARGV
directories.each do |d|

  Find.find("#{d}") do |file|
    if file.match(pattern)
      if File.owned?("#{file}")
         puts "Deleting #{file}"
   File.delete("#{file}")
      end
    end

  end
end

The default pattern is set to delete backup files ending with ”~”. Below is the ouput:

1
2
3
4

dev@testing$(~) ./rmbackups.rb
Deleting /home/dev/ruby/findit.rb~
Deleting /home/dev/ruby/delicious.rb~
Deleting /home/dev/ruby/rmbackups.rb~

You might change the pattern from the command line:

1
2
3
4
5

.dev@testing$(~) /rmbackups.rb /var/tmp /tmp  ".?#.*#$"
Deleting /var/tmp/#.vile.somethin#
Deleting /var/tmp/.#strangefile#
Deleting /tmp/#file#
Deleting /tmp/#a-strangefile#

This is how it works, I did not try too many different patterns beyond these.

1.CDPATH

If this is set, keeps a directory list separated by ‘:’ and every time that you type the command “cd DIR” it searches for the directory in the above list, even if your current directory is not the right one. Let’s see an example, I define the directory holding all my git repositories, if I try to make cd repo from any part on the system, even if I not in the right place, I will get there:

1
2
3
4
5

export CDPATH="~/mygits"
devadm@testing$(~)ls mygits
blog scripts org-mode cv
devadm@testing$(/usr/share/doc) cd  blog
devadm@testing$(~/mygits/blog)

In my opinion is useful to keep just one directory or probably up to two, more than this, it could get messy.

2.FIGNORE

If you want to narrow down the ouput when performing filename completion, this is without a doubt your shell env. When set, it ignores suffixes while performing filename completion.

1
2
3
4
5

devadm@testing$(~)export FIGIGNORE="#:.o:~"
devadm@testing$(~/application)ls file*
file1.o file2.o file1 file1 file~
devadm@testing$(~)ls [TAB] [TAB]
file1 file2

I mentioned this shell env because it was curious to me. I guess I have never found any case where I had to use such a thing, but that does not mean it wouldn’t be useful to somebody else.

3.HISTCONTROL

It can be set to ignorespace , and it does not record words starting with empty space. On the other hand I do not like to have duplicated lines, ingnoredups does not repeat entries in the history. Finally, I like to join both options so I use ignoreboth instead.

1

export HISTCONTROL=ignoreboth

4.HOSTFILE

It holds the path to a file containing a list of hosts in the same format as /etc/hosts , if it is set, tries to complete the hostname with one of the entries on the file. Otherwiese will look for /etc/hosts.

It could be pretty useful if you want to keep a personal file with your hosts.

1

export HOSTFILE="~/.myhosts"

The only problem I see here, if I you want to access hosts defined in /etc/hosts. You could make a script that checks if there is a new entry in the /etc/hosts and then appends the last entries to your personal list. Once again, this is only an idea I came up with, but actually I did not follow through.

5.TMPDIR

If set, bash uses its value as the name of a directory in which creates temporary files. With this option I can set my personal temp directory.

1
2

 export TMPDIR="~/.tmpbash"

6.TMOUT

Sets a timeout and affects to read or select builtin commands, when not input is given. An interesting case is if you set this to your current shell. After N seconds without providing any input it will kill your shell, so keep this in mind.

1
2
3
4
5
6
7
8
9

#!/bin/bash

# Sets to 3 seconds timeout
TMOUT="3"

printf 'Could you please give me an absolute path ?'
read -r -s -n10 absolute_path

[ -z name ] && printf 'Your path is : %s\n' $absolute_path

Well, the last posts were focused on bash shell environment because I was digging into the man bash and I found quite interesting things I did not know, I hope some of them were useful for you as well.

1. BASHPID

This is the PID of the current bash process. This behaviour is different from $$ in cases such a subshell where $BASHPID says the PID of the subshell, whereas $$ shows the PID of the bash holding the subshell.

1
2

$ echo $$ $BASHPID # 23353 23353
$ echo 'Subsshell' $(echo $$; echo $BASHPID)# 21060 23353

2. BASH_LINENO

Number of lines in the current script, from the beginning to the line the function was called from.

1
2
3
4
5
6
7
8
9
10
11

#!/bin/bash 

function show_env_vars()

  echo "$FUNCNAME"    # show_env_vars
  echo "$BASH_LINENO" # Number on lines till show_env_vars was call (11)
  echo "$LINENO"      # Current line 7

}

show_env_vars

3. DIRSTACK

Array with the directories you are moving using popd and pushd builtins to add/remove directories.

1
2
3
4
5
6
7
8
9
10

# After doing some pushd 

$ for i in  ${DIRSTACK[@]}; do printf  'DIR: %s\n' $i; done                                                                                                                                                               

DIR: /tmp
DIR: /var/tmp/testing
DIR: /home/cartoon
DIR: /usr/share/doc
DIR: /var/tmp
DIR: /var/www

4. EUID and GROUPS

EUID of the current user, initialized at shell startup. This variable is read only. On the other hand GROUPS is an array of groups which the current user is member of.

1
2
3

for ((i=0; i < ${#GROUPS[@]};  i++)); do printf 'gid:%d\n' ${GROUPS[$i]}" ; done
gid: 1002
gid: 1001

5. Gathering OS info

Below there are some of the shell environment vars you could use, for instance if you need to check the architecture. I have not tried in Sparc yet but I would like.

1
2
3
4

HOSTNAME='Self explanatory'
HOSTTYPE='Arhcitecture i486'.
OSTYPE='Operating System where bash is running i.e: linux-gnu'
PPID='Parent process ID of the current shell'

6. Random numbers

1
2
3
4

RANDOM A random number betwwen 0 and 32767.
# Getting a random number between 0 and 99

echo "RANDOM: $(( $RANDOM % 100))

In sum, there are a wide number of shell environment variables to keep in mind if you are scripting. Next time I will post more shell vars, but focused on customizing your bash shell.

1.BASH_ENV

If you run a script, it looks for this variable and if it’s set, expands the value to the name of a file and reads its content. The filename should content the absolute pathmane otherwise it will no be able to locate the file.

1

export BASH_ENV="$HOME/.customs"

Take a look to the above example. From now, and beacuse I set BASH_ENV in my .bashrc, my scripts will have a set of fuctions or whatever I defined in there. It would be the same doing source $HOME/.customs in every script. This is really good, you have a set of custom functions or variables available to all your scripts.

A good idea would be to set this variable in either /etc/profile or /etc/bash.bashrc , if you want to share it with the rest of users in the system.

2.Run in a subshell ( ) vs current shell{ }

This is someting you might not pay attention. In order to assign the ouput of some shell commands to a variable, try first to use { }. Let’s see an example:

1
2
3
4
5
6
7
8
9

# Current shell
declare -A colors=();

line=''
{ for key in ${!colors[@]}; do line+="$key" ; done; }

# Subshell
line=
$(for key in ${!colors[@]}; do line+="$key" ; done;)

Perhaps you realized the problem it crops up here.

Current shell {}

-Variables are available while the script is running. In the previus example I can access both variables, colors and line. Commands are separated by ‘;’. It’s also quite handy when using conditionals.

1

 if_true  &&  { cmd1; cmd2; cmd3; }

Subshell $()

-Variable assigments do not remain. Any change to any variable in the script would not take effect after ending the command execution. However it would be possible to get the output by means of either echo or printf commands. Besides variable scope, performance could be worse due to new subshell execution.

3.Using [[ ]] expressions

You might know the old form [ ] , but this is the new one, and has some pretty cool properties as Pattern Matching. I guess the best way to get an idea is watching an example.

1
2
3
4
5
6
7
8
9
10

ip_address="10.20.30.40"
[[ $ip_address =~ ^([0-9]{1,3})\.([0-9]{1,3})\.([0-9]{1,3})\.([0-9]{1,3})$ ]] && echo "Valid IP address"

# Iterating matches
for n in ${BASH_REMATCH[@]}; do echo $n; done
10.20.30.40
10
20
30
40

Of course the above example is just for purpose, actually it does not validate a real ip address, but does the trick. Most interesting is the variable BASH_REMATCH, an array holding each substring matched by parenthesized subexpressions. Regarding regular expressions you should look for in the man as in regex(3) and regex(7).

4. Case and ;& operator

This operator continues the execution to next option if available. I came up with an example and you will see how it works:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39

options="ac:lh"
while getopts $options  flag
do
        case $flag in
                a)
                        list_all="on"
                        ;&
                c)

                        c=${OPTARG,,}
                        [ $list_all == "on" ]  && shift
                        [ $list_all == "off" ] && shift && shift

                        args="$*"
                        msg="${args:-$default}"


                        if [ $list_all == "on" ];then
                                for color in ${!colors[@]}
                                do
                                        draw_with_color "$color" "$msg"
                                done
                        else
                                check_color "$c" && draw_with_color "$c" "$msg"
                        fi

                           ;;
                l|-list)
                        printf 'Available colors: \n'
                        _show_colors && exit
                        ;;
                h|-help)
                        usage && exit
                        ;;
                *)
                        usage && exit
                        ;;
        esac
done

Previus chunk of code is part of the pickcolor script. The main idea was to jazz some text up with a chosen color, then I thought it would be more practical to enable the option of pating the same text with all the available colors.

1
2
3
4
5
6
7
8
9
10
11
12
13
14

pickcolor.sh
usage: pickcolor.sh [OPTIONS] message
Options:
       -a --all    Test all colors for message
       -c --color  Set font color.
       -l --list   List available colors.
       -h --help   Help

# Colors up using all the colors
$ pickcolor -a  Show me the colors

# Color up using just one color
$ pickcolor -c red Paint my room !!

The special operator ;& was really useful and did the trick. Next part I will write about shell vars and some builtin commands to have in mind.

In my daily work I have to fill in many templates in order to configure some network elements such as switches or routers. Likewise, I decided to make a small vim script that will help me out in such an unpleasant task.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16

function! Net_acl()
  let s:vip_a  = input('VIP-A:  ')
  let s:vip_b  = input('VIP-B:: ')
  let s:vip_c  = input('VIP-C:  ')
  let s:vip_d  = input('VIP-D:  ')

  :%s/<vlab-a>/\=s:vip_a/g
  :%s/<vlan-b>/\=s:vip_b/g
  :%s/<vlan-c>/\=s:vip_c/g
  :%s/<vlan-dP>/\=s:vip_d/g
endfunction

function! Net_vlan_names()
  :%s/<vlan-\(\w*\)>/vlan-\1/g
  :%s/<\(extravlan\)>/\1/g
endfunction

When the function Net_acl() is called, it will ask the ips addresses you’d like to use for each of the vlans you have defined. Keep in mind, when you define a function in VIM, the first letter must be capital.

1

:call Net_acl()

On the other hand, the function Net_vlan_names() takes a pattern inside angle brackets off. If you have to do repetitive tasks, scripting is the right approach no matter what you use. Just get the job done.

From now on, I will keep on adding more functions to my helpers.vim. I really enjoyed.

Then I decided it would be a great opportunity to try Puppet, and I’ve made a couple of recipes, just for users and openssh service. I will write someday about this, right now I want to explain some issues I came across whereas I peformed the migration of this blog and my wiki.

I had some issues installing Nginx + PHP-fpm. I use Dokuwiki to keep my technical information. You might install Nginx either from Debian repositories or sources. Due to this site runs using Toto engine, and I need Passenger, I asked a friend of mine which would be the best way to do such a thing, and he recommended me to install from Ruby Enterprise.

The fastcgi_params file is different depending on how you install Nginx. Installing from Passenger installer, this file does not have the follwing line.

fastcgi_param   SCRIPT_FILENAME         $request_filename

After adding the above line I got properly working Nginx + php-fpm.

UPDATED:Another issue I have just seen, Nginx init script does not stop the processes. You have to be sure the _pid /var/run/nginx.pid __ parameter is defined in the nginx.conf file, otherwise the processes will not be killed.

 ERROR (SrvDaemon:349) Exception starting xend ((38, 'Function not implemented'))
Traceback (most recent call last):
  File "usr/lib/python2.5/site-packages/xen/xend/server/SrvDaemon.py", line 335, in run
    xinfo = xc.xeninfo()<

The function xeninfo() is only available in Xen 4.x , that’s what I found out searching in Google. The problem, I messed everything up. The way I sorted it out was pretty simple, just uninstall all the packages and reinstall them, at least in Debian Lenny it worked well.

That’s all.

./myapp: error while loading shared libraries: libtest.so: cannot open shared object file: No such file or direct

Let’s take a look inside the binary:

tuxedo@host:$> ldd test
    linux-gate.so.1 =>  (0xb7ef6000)
    libatest.so => not found
    libc.so.6 => /lib/i686/cmov/libc.so.6 (0xb7d81000)

By default the system is looking for in the paths defined in the /etc/ld.so.conf, which recursively adds the definitions in the folder /etc/ld.so.conf.d Here’s the quick trick:

tuxedo@host:$> export LD_LIBRARY_PATH=`pwd` 

I use this whereas I’m implementing my library, after that you can put it wherever you feel like.

tuxedo@host:$> ldd test
linux-gate.so.1 =>  (0xb7f5e000)
    libtest.so => /home/tuxedo/syslib/libtest.so (0xb7f56000)
    libc.so.6 => /lib/i686/cmov/libc.so.6 (0xb7de3000)
    /lib/ld-linux.so.2 (0xb7f5f000)

Now the executable will work. I’ll retake this issue, I have some interesting things to tell about shared libraries. Happy coding!!

 # aptitude install x11-server-utils

If you executes xrandr, it will show you the outputs you have available:

$ xrandr VGA connected (normal left inverted right)
 1280x960       60.0
   1280x800       60.0  
   1152x768       54.8  
   800x600        56.2  
   640x480        59.9  
LVDS connected 1280x800+0+0 (normal left inverted right) 0mm x 0mm
   1280x800       59.9*+   60.0  
   1280x768       60.0  
   1152x768       54.8  
   1024x768       85.0     75.0     70.1     60.0  
   832x624        74.6  
   800x600        85.1     72.2     75.0     60.3     56.2  
   640x480        85.0     72.8     75.0     59.9  
   720x400        85.0  
   640x400        85.1  
   640x350        85.1  

To get all this working, a section must be added into /etc/X11/xorg.conf file

Section "Screen"
    Identifier      "Default Screen"
    Monitor         "Configured Monitor"
    DefaultDepth    24
    <strong>SubSection "Display"
        Modes "1280x1024" "1280x800"
        Virtual 2560  1024
    EndSubSection</strong>
EndSection

Using xrandr

Let’s to try out our configuration:

$ xrandr --output VGA --mode 1280x1024 --left-of LVDS

In fact, that is the line we need to put in our little script:

$ vi .kde/Autostart/dual.sh

Henceforth we can work with an extended desktop, and increase our productivity.

3. LDAP setup

Let’s guess your root suffix is dc=company,dc=com, you need to append the next entry to your directory :

dn: ou=sudoers,dc=company,dc=com
  objectClass: top
  objectClass: organizationalunit
  description: Sudo Configuration
  ou: sudoers

Besides we will need a default profile:

dn: cn=defaults,ou=sudoers,dc=company,dc=com
  sudoOption: ignore_local_sudoers
  objectClass: top
  objectClass: sudoRole
  cn: defaults
  description: Our default options
  sudooption: log_host
  sudooption: logfile=/var/log/sudolog
  sudooption: !syslog

Perhaps you would like to get the most of sudo’s powder, take a look in its website. You can add as much profiles as you like, suppose you want to add one for system administration:

dn: cn=sysadmin,ou=sudoers,dc=company,dc=com
 objectClass: top
 objectClass: sudoRole
 cn: unix_admins
 sudoUser: tuxman
 sudoUser: darkman
 sudoUser: bill
 sudoHost: ALL
 sudoCommand: /usr/bin/ls

As far as I concern, how to configure sudo is out of this post, however together the source of sudo there is an utility, _sudoers2ldif__, a perl script that helps you to translate your sudo’s configuration file. Next step requires to modify our profile. Probably you will have a similar profile to this one:

dn: cn=default,ou=profile,dc=company,dc=com
objectClass: DUAConfigProfile
defaultSearchBase: dc=company,dc=com
cn: default
credentialLevel: proxy
defaultServerList: 192.168.76.66
profileTTL: 300
searchTimeLimit: 60
authenticationMethod: simple
serviceSearchDescriptor: passwd:cn=sudoers,dc=company,dc=com

After these modifications you must initialize your client.

4. Setting up /etc/ldap.conf and nsswitch.conf

It’s time to tell our client where to find sudoers file, by means of /etc/ldap.conf, that looks something like this.

   uri ldap://192.168.76.66
   sudoers_base ou=sudoers,dc=company,dc=com
   bindpw  cn=proxyagent,ou=profile,dc=company,dc=com
   binddn  password
   sudoers_debug 0

You might use anonymous access, that’s your choice, just remember to check your ACI’s. Pretty interesting the option sudoers_debug which helps you to debug, at level 3 will show you as much information as possible. The last step, how to find our sudoers’ profile, nsswitch.conf

  sudoers: ldap

Let’s check if is working:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31

tuxman@host:$> sudo ls
[sudo] password for client:

sudo ls
LDAP Config Summary
===================
uri              ldap://192.168.76.66
ldap_version     3
sudoers_base     ou=sudoers,dc=company,dc=com
binddn           (anonymous)
bindpw           (anonymous)
ssl              (no)
===================
sudo: ldap_initialize(ld, ldap://192.168.76.66)
sudo: ldap_set_option: debug -> 0
sudo: ldap_set_option: ldap_version -> 3
sudo: ldap_sasl_bind_s() ok
sudo: found:cn=defaults,ou=sudoers,dc=company,dc=com
sudo: ldap sudoOption: 'ignore_local_sudoers'
sudo: ldap sudoOption: 'log_host'
sudo: ldap sudoOption: 'logfile=/var/log/sudolog'
sudo: ldap sudoOption: '!syslog'
sudo: ldap search '(|(sudoUser=tuxman)(sudoUser=%other)(sudoUser=ALL))'
sudo: found:cn=sysadmin,ou=sudoers,dc=company,dc=com
sudo: ldap sudoHost 'ALL' ... MATCH!
sudo: ldap sudoCommand '/usr/bin/ls' ... MATCH!
sudo: Command allowed
sudo: user_matches=1
sudo: host_matches=1
sudo: sudo_ldap_lookup(0)=0x02
tuxman@host:$> files/  sudoers2ldif.pl

At this point everything should be working. Last step, to translate our sudoers file.

The first one gets a backup of our database, the second rotates the files. I also add support to Syslog, because I would like to know if my script worked out.

1
2
3
4

#!/bin/bash
DATE=$(date +%Y-%m-%d)
mysqldump -h host userdb  database  --password=1234 | gzip > \
 ${HOME}/path/files/file-${DATE}.gz

Crontab

Let’s add some entries to our crontab:

user@home:~> crontab -e

As far as I concern I would review the entries, it’s pretty easy to make some mistakes while we’re writing, besides I guess it’s a best practice.

user@home:~> crontab -l

The above command will show our entries.

1
2
3
4
5
6

# m h  dom mon dow   command
HOME=/path/user/home
PATH=PATH:/path/to/ruby

0 00 * * * ${HOME}/scripts/backup_db_blog.sh
0 00 * * * ${HOME}/scripts/rotatedb.rb

It’s a MUST to leave an empty line at the end of the cron, otherwise the cron will not run. There is another thing to take into account, PATHS. Don’t forget to set them.

Syslog

Due to I wanted support Syslog, I had to set it up correctly in my /etc/syslog.conf

local7.*                  /var/log/backups.log

Aside to create the above file, you must reload syslog daemon. Afterwards, I wanted to know if that configuration would work. There is a command that will help you:

user@home:~> logger -p local7.debug "Sending a message to debug"
user@home:~> more /var/log/backups.log
Aug  6 20:31:22 home user: Sending a message to debug

The first argument local7 is the Syslog’s FACILITY and the other one is the PRIORITY. You must adapt to your own script.
Here is the script to rotate. After then days, will remove only the four oldest backups.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29

#!/usr/bin/env ruby
%w(syslog).each {|c| require c }

BACKUP='/path/to/backup/directory'

module SyslogMsg
  Syslog.open("rotatedb", Syslog::LOG_PID  \
             | Syslog::LOG_CONS ,  Syslog::LOG_LOCAL7 )
  def send(msg="Message sent")
    Syslog.log(Syslog::LOG_DEBUG, msg)
  end
end

include SyslogMsg
files = []

Dir.entries(BACKUP).each do |e|
  if e !~ /^\./
    files << e
  end
end
files.sort!

if files.length > 10
  0.upto(4) do |index|
    File.delete( BACKUP + "/" + files[index] )
  end
  SyslogMsg::send("Backups for mysql rotated.")
end

Finally, if you get some troubles, review the above steps, check that you have the right permissions and you have reloaded your syslog configuration. If you try the logger command and you see the message you have just sent, everything should work out.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

tuxman@athome:> ssh -p 3000 someuser@localhost

@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
@    WARNING: REMOTE HOST IDENTIFICATION HAS CHANGED!     @
@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@</b>
IT IS POSSIBLE THAT SOMEONE IS DOING SOMETHING NASTY!
Someone could be eavesdropping on you right now (man-in-the-middle attack)!
It is also possible that the RSA host key has just been changed.
The fingerprint for the RSA key sent by the remote host is
22:ce:a2:e1:fe:cc:e6:73:cb:03:96:1e:23:3c:5b:55.
Please contact your system administrator.
Add correct host key in /home/tuxman/.ssh/known_hosts to get rid of this message.
Offending key in /home/tuxman/.ssh/known_hosts:159
RSA host key for [localhost]:4000 has changed and you have requested strict checking.
Host key verification failed.

Appending some lines to the ~/.ssh/config file, solves the issue

Host localhost
   HostKeyAlias  <em>youralias</em>

The HostKeyAlias keyword allows us to define an alias and use it instead of the real name. I frequently use ssh port forwarding and I wanted to get rid of this message.

Moreover, bear in mind you would have to update every sudo config file , definitely would be tedious, and here’s when LDAP gets in.

I brought into play two virtual machines, both of them running Solaris 10. Commonly I prefer doing that before making some huge mistake in a real environment, so here’s what I did. I called box0 to the client and ldapbox to the server. The rest of the post, I’ll assume you have set up a Directory Server and Native LDAP client service working fine. Summary:

1. Extending Schema

This step is pretty straightforward, you only need to add the schema to your directory instance, and restart the server. Assuming your instance might be in the default path /var/opt/SUNWdsee/dsins1/config/schema/99users.ldif.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

attributeTypes: ( 1.3.6.1.4.1.15953.9.1.1 NAME 'sudoUser' DESC 'User(s) who may run sudo'
 EQUALITY caseExactIA5Match SUBSTR caseExactIA5SubstringsMatch SYNTAX
 1.3.6.1.4.1.1466.115.121.1.26 X-ORIGIN 'SUDO' )

attributeTypes: ( 1.3.6.1.4.1.15953.9.1.2 NAME 'sudoHost' DESC 'Host(s) who may run sudo' 
EQUALITY caseExactIA5Match SUBSTR caseExactIA5SubstringsMatch SYNTAX
1.3.6.1.4.1.1466.115.121.1.26 X-ORIGIN 'SUDO' )

attributeTypes: ( 1.3.6.1.4.1.15953.9.1.3 NAME 'sudoCommand' DESC' Command(s) to be executed by sudo' 
EQUALITY caseExactIA5Match SYNTAX 1.3.6.1.4.1.1466.115.121.1.26
X-ORIGIN 'SUDO' )

attributeTypes: ( 1.3.6.1.4.1.15953.9.1.4 NAME 'sudoRunAs' DESC 'User(s) impersonated by sudo' 
EQUALITY caseExactIA5Match SYNTAX 1.3.6.1.4.1.1466.115.121.1.26
X-ORIGIN 'SUDO' )

attributeTypes: ( 1.3.6.1.4.1.15953.9.1.5 NAME 'sudoOption' DESC 'Options(s) followed by sudo'
 EQUALITY caseExactIA5Match SYNTAX 1.3.6.1.4.1.1466.115.121.1.26
 X-ORIGIN 'SUDO' )

objectClasses: ( 1.3.6.1.4.1.15953.9.2.1 NAME 'sudoRole' SUP top
STRUCTURAL DESC 'Sudoer Entries' MUST ( cn ) MAY ( sudoUser $ sudoHost
$ sudoCommand $ sudoRunAs $ sudoOption $ description ) X-ORIGIN 'SUDO'
)

Now you only need to restart the server.Why do I need to restart my server? The matter is, the first time you started your server, the file was read into memory, so any change you make later will not have effect, at least you restart the instance.

2. LDAP support for sudo

You will need to get the source code of sudo and at least 1.7 version or upper. The reason is because earlier versions will not read nsswitch.conf. I used sudo-1.7.2p7, you can get it from Sunfreeware. Of course if you want to compile you will need to solve some dependencies, here is the list, however you had better confirm by yourself.

1
2
3
4
5
6
7
8
9
10
11
12

/---|(Sudo)
       |-* gcc-3.4.6-sol10-x86-local  
       |-* libiconv-1.13.1-sol10-x86-local
       |-* libintl-3.4.0-sol10-x86-local 
       |-* openssl-1.0.0a-sol10-x86-local
  

 /---|( OpenLdap )
       |- * db-4.7.25.NC-sol10-x86-local
       |- * libtool-2.2.6b-sol10-x86-local
       |- * sasl-2.1.21-sol10-x86-local
       |- * openldap-2.4.22-sol10-x86-local

At this point and after installing all the dependencies we just need to compile:

./configure --with-ldap && make 

Those people who like tinkering with Unix tools, is time to call ldd and take a look into sudo.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

libpam.so.1 =>   /lib/libpam.so.1
        libdl.so.1 =>    /lib/libdl.so.1
*       libldap-2.4.so.2 =>      <em><b>(/usr/local/lib/libldap-2.4.so.2)</b></em>
*       liblber-2.4.so.2 =>      <em><b>(/usr/local/lib/liblber-2.4.so.2)</b></em>
        libintl.so.8 =>  /usr/local/lib/libintl.so.8
        libsocket.so.1 =>        /lib/libsocket.so.1
        libnsl.so.1 =>   /lib/libnsl.so.1
        libc.so.1 =>     /lib/libc.so.1
        libcmd.so.1 =>   /lib/libcmd.so.1
        libresolv.so.2 =>        /usr/lib/libresolv.so.2
        libgen.so.1 =>   /usr/lib/libgen.so.1
        libsasl2.so.2 =>         /usr/local/lib/libsasl2.so.2
        libssl.so.1.0.0 =>       /usr/local/ssl/lib/libssl.so.1.0.0
        libcrypto.so.1.0.0 =>    /usr/local/ssl/lib/libcrypto.so.1.0.0
        libgcc_s.so.1 =>         /usr/local/lib/libgcc_s.so.1
        libiconv.so.2 =>         /usr/local/lib/libiconv.so.2
        libsec.so.1 =>   /usr/lib/libsec.so.1
        libmp.so.2 =>    /lib/libmp.so.2
        libmd.so.1 =>    /lib/libmd.so.1
        libscf.so.1 =>   /lib/libscf.so.1
        libavl.so.1 =>   /lib/libavl.so.1
        libdoor.so.1 =>  /lib/libdoor.so.1
        libuutil.so.1 =>         /lib/libuutil.so.1
        libm.so.2 =>     /lib/libm.so.2

If the LDAP libraries does not appear remember to add the path:

# crle -l -u  PATH_TO_LIBRARIES</b>

Obviously I wouldn’t like having to install OpenLdap in all my clients (if you want to apply to more than one), so I thought to carry just with the libraries I needed. In a nutshell, we have just extended the schema and also enabled ldap support for sudo. The two last points for the next post.

Due to my lack of memory I decided to make a function that shows me how many programs I have in background, here is the function:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20

function get_njobs {
 njobs=$(jobs | wc -l)

 if [ $njobs -gt "0" ];
 then
   echo $njobs |  sed -e 's/\([0-9]*\)/(\1)/g'
 fi
}

function prompt {

#  Shell into Emacs
 if [ $TERM != "dumb" ];
 then
  alias ls='ls --color=auto'
  PS1="${purpple}\u@${close}${YELLOW}\h${close}:{\W}\$(get_njobs)"
 else
        PS1="[\u@\h:\w]"
 fi
}

Now you can do some work in background and will see what I mean:

user@host:{~} emacs &
user@host:{~}(1)

As you may notice, the number inside the brackets remembers me if there is some program running in background. Do not forget to add the function at the end of your bashrc.

UPDATE: If you add \j to your PS1 variable you will get the same effect but if there is not jobs you will always get a (0).